N-nitroso compounds are environmentally encountered by humans in occupational settings, during food consumption and from cigarette smoke. Many N-nitroso compounds generate mutations via alkylation damage of DNA. All cells synthesize a protein that specifically and directly repairs the most potent mutagenic lesion produced by monofunctional alkylating agents, namely 06-alkylguanine (06alkylGua). Activity levels of the repair protein, O6-methylguanine-DNA methyltransferase (MGMT), are low in the brain compared to the liver. An inverse correlation between tumor formation subsequent to DNA alkylation with the level of MGMT activity has been inferred. Brain tissue, which displays low levels of MGMT activity, is highly susceptible to tumor formation after DNA alkylation damage. Primary brain tumors result in more deaths than Hodgkin's disease or multiple sclerosis in adults and brain tumors may be increasing in the elderly population. The current prognosis for patients with brain tumors is dismal. The proposed project is designed to more clearly define the role of MGMT in carcinogenesis and to characterize modes of regulation of MGMT gene expression. The hypothesis that the frequency of alkylation carcinogenesis is inversely proportional to the MGMT activity level for a given tissue will be tested. Because of the relationship between brain tumor formation after alkylation damage and MGMT activity, a portion of the proposal involves generating transgenic mice that overexpress human MGMT (hMGMT) in brain tissue. High levels of MGMT expression in brain is assured by using the human transferrin promoter (hTF) which has demonstrated high expression in brain tissue. hTF/MGMT transgenic mice will be exposed to DNA alkylating agents to analyze the effects of increased MGMT activity on tumorigenesis. As a long term goal to determine the effects of reduced MGMT activity on tumor formation after mutagen exposure, homologous recombination techniques will be employed with embryonic stem (ES) cells to disrupt one copy of endogenous murine MGMT. Disruption or alteration of one copy of MGMT should effectively reduce MGMT activity. The molecular mechanism(s) that result in low MGMT activity in brain relative to liver remain undefined. The hypothesis that tissue-specific cis-acting DNA regulatory elements function to produce low MGMT activity in brain and high MGMT activity in liver will be tested. To our knowledge, mechanisms of transcriptional regulation for a mammalian DNA repair gene have not yet been described. Circumstances are now propitious for identifying cis-acting DNA regulatory elements that may determine the levels of MGMT activity in various tissues. These approaches will facilitate progress toward defining the role of DNA repair genes in alkylation-induced tumor formation.